Top Banner
7/21/2019 Anova w High Speed Trial http://slidepdf.com/reader/full/anova-w-high-speed-trial 1/29 General Linear Model [ Dat aSet7] J : \ Bi gRoomExperi ment s - J an- 22- 2014\ Ri t aLBOBr ad\ RMdata_meanMi dSl i c eW i dt h. sav Within-Subjects Factors Measure: MeanMidsliceWidth trial Dependent Variable 1 2 3 Pre Post1 trial3_fast Between-Subjects Factors N MovieCond .000000 1.000000 2.000000 55 63 57 Multivariate Tests a Effect Value F Hypothesis df Error df Sig. trial Pillai's Trace Wilks' Lambda Hotelling's Trace Roy's Largest Root trial * MovieCond Pillai's Trace Wilks' Lambda Hotelling's Trace Roy's Largest Root .164 16.736 b 2.000 171.000 .000 .836 16.736 b 2.000 171.000 .000 .196 16.736 b 2.000 171.000 .000 .196 16.736 b 2.000 171.000 .000 .028 1.233 4.000 344.000 .297 .972 1.234 b 4.000 342.000 .296 .029 1.236 4.000 340.000 .295 .029 2.478 c 2.000 172.000 .087 Design: Intercept + MovieCond Within Subjects Design: trial a. Exact statistic b. The statistic is an upper bound on F that yields a lower bound on the significance level. c. Page 1
29

Anova w High Speed Trial

Mar 04, 2016

Download

Documents

Brian Finch

Describibg ANOVAs at High Speed
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 1/29

General Linear Model

[ DataSet7] J : \ Bi gRoomExper i ment s - J an- 22- 2014\ Ri t aLBOBr ad\ RMdata_meanMi dSl i c

eWi dt h. sav

Within-SubjectsFactors

Measure:

MeanMidsliceWidth

trial

Dependent

Variable

1

2

3

Pre

Post1

trial3_fast

Between-Subjects Factors

NMovieCond .000000

1.000000

2.000000

55

63

57

Multivariate Testsa

Effect Value F Hypothesis df Error df Sig.

trial Pillai's Trace

Wilks' Lambda

Hotelling's Trace

Roy's Largest Root

trial * MovieCond Pillai's Trace

Wilks' Lambda

Hotelling's Trace

Roy's Largest Root

.164 16.736b

2.000 171.000 .000

.836 16.736b

2.000 171.000 .000

.196 16.736b

2.000 171.000 .000

.196 16.736b

2.000 171.000 .000

.028 1.233 4.000 344.000 .297

.972 1.234b

4.000 342.000 .296

.029 1.236 4.000 340.000 .295

.029 2.478c

2.000 172.000 .087

Design: Intercept + MovieCond

Within Subjects Design: trial

a.

Exact statisticb.

The statistic is an upper bound on F that yields a lower bound on the significance level.c.

Page 1

Page 2: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 2/29

Mauchly's Test of Sphericitya

Measure: MeanMidsliceWidth

Within Subjects Effect Mauchly's W

 Approx. Chi-

Square df Sig.

Epsilonb

Greenhouse-

Geisser trial .964 6.311 2 .043 .965

Mauchly's Test of Sphericitya

Measure: MeanMidsliceWidth

Within Subjects Effect

Epsilonb

Huynh-Feldt Lower-bound

trial .987 .500

Tests the null hypothesis that the error covariance matrix of the orthonormalized transformed dependent variables

is proportional to an identity matrix.

Design: Intercept + MovieCond

Within Subjects Design: trial

a.

May be used to adjust the degrees of freedom for the averaged tests of significance. Corrected tests are

displayed in the Tests of Within-Subjects Effects table.

b.

Tests of Within-Subjects Effects

Measure: MeanMidsliceWidth

Source

Type III Sum

of Squares df Mean Square F

trial Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

trial * MovieCond Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

Error(trial) Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

103.564 2 51.782 17.359

103.564 1.930 53.658 17.359

103.564 1.974 52.457 17.359

103.564 1.000 103.564 17.359

16.045 4 4.011 1.345

16.045 3.860 4.157 1.345

16.045 3.949 4.063 1.345

16.045 2.000 8.022 1.345

1026.123 344 2.983

1026.123 331.971 3.091

1026.123 339.571 3.022

1026.123 172.000 5.966

Page 2

Page 3: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 3/29

Tests of Within-Subjects Effects

Measure: MeanMidsliceWidth

Source Sig.

trial Sphericity AssumedGreenhouse-Geisser 

Huynh-Feldt

Lower-bound

trial * MovieCond Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

Error(trial) Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

.000

.000

.000

.000

.253

.254

.253

.263

Tests of Within-Subjects Contrasts

Measure: MeanMidsliceWidth

Source trial

Type III Sum

of Squares df Mean Square F Sig.

trial Linear  

Quadratic

trial * MovieCond Linear 

Quadratic

Error(trial) Linear  

Quadratic

82.115 1 82.115 23.266 .000

21.449 1 21.449 8.804 .003

10.453 2 5.226 1.481 .230

5.592 2 2.796 1.148 .320

607.059 172 3.529

419.063 172 2.436

Tests of Between-Subjects Effects

Measure: MeanMidsliceWidth

Transformed Variable: Average

Source

Type III Sum

of Squares df Mean Square F Sig.

Intercept

MovieCond

Error 

19501.650 1 19501.650 1832.483 .000

7.747 2 3.873 .364 .695

1830.459 172 10.642

Estimated Marginal Means

Page 3

Page 4: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 4/29

1. MovieCond

Measure: MeanMidsliceWidth

MovieCond Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

.000000

1.000000

2.000000

5.932 .254 5.430 6.433

6.170 .237 5.701 6.638

6.213 .249 5.721 6.706

2. trial

Measure: MeanMidsliceWidth

trial Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

1

2

3

6.733 .189 6.360 7.107

5.819 .160 5.503 6.134

5.763 .184 5.400 6.126

3. MovieCond * t rial

Measure: MeanMidsliceWidth

MovieCond trial Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

.000000 1

2

3

1.000000 1

23

2.000000 1

2

3

6.682 .337 6.017 7.347

5.847 .285 5.285 6.410

5.265 .328 4.619 5.912

6.672 .315 6.051 7.293

5.740 .266 5.214 6.2656.097 .306 5.493 6.701

6.845 .331 6.192 7.498

5.869 .280 5.316 6.421

5.926 .322 5.291 6.561

Profile Plots

Page 4

Page 5: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 5/29

trial

321

   E  s   t   i  m  a   t  e   d   M  a  r  g   i  n  a   l   M  e  a  n  s

7.00

6.50

6.00

5.50

5.00

Estimated Marginal Means of MeanMidsliceWidth

2.000000

1.000000

.000000

MovieCond

DATASET ACTI VATE Dat aSet 7.

SAVE OUTFI LE=' J : \ Bi gRoomExper i ment s - J an- 22- 2014\ Ri t aLBOBr ad\ RMdat a_meanMi dS

l i ceWi dt h. sav'

  / COMPRESSED.

DATASET ACTI VATE Dat aSet 2.

DATASET CLOSE Dat aSet 7.

DATASET ACTI VATE Dat aSet 8.

GLM Pre Post 1 t r i al 3_f ast BY Movi eCond

  / WSFACTOR=t r i al 3 Pol ynomi al

  / MEASURE=sd_mi dsl i cewi dth

  / METHOD=SSTYPE( 3)

  / PLOT=PROFI LE( t r i al *Movi eCond)

  / EMMEANS=TABLES( Movi eCond)

  / EMMEANS=TABLES( t r i al )

  / EMMEANS=TABLES( Movi eCond*t r i al )

  / CRI TERI A=ALPHA( . 05)

Page 5

Page 6: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 6/29

  / WSDESI GN=t r i al

  / DESI GN=Movi eCond.

General Linear Model

[ DataSet8] J : \ Bi gRoomExper i ment s - J an- 22- 2014\ Ri t aLBOBr ad\ RMdata_sdmi dsl cwdth. sav

Within-Subjects

Factors

Measure:

sd_midslicewidth

trial

Dependent

Variable

1

2

3

Pre

Post1

trial3_fast

Between-Subjects Factors

N

MovieCond .000000

1.000000

2.000000

55

63

57

Multivariate Testsa

Effect Value F Hypothesis df Error df Sig.

trial Pillai's Trace

Wilks' Lambda

Hotelling's Trace

Roy's Largest Root

trial * MovieCond Pillai's Trace

Wilks' Lambda

Hotelling's Trace

Roy's Largest Root

.092 8.624b

2.000 171.000 .000

.908 8.624b 2.000 171.000 .000

.101 8.624b

2.000 171.000 .000

.101 8.624b

2.000 171.000 .000

.035 1.527 4.000 344.000 .194

.965 1.527b

4.000 342.000 .194

.036 1.527 4.000 340.000 .194

.032 2.789c

2.000 172.000 .064

Design: Intercept + MovieCond

Within Subjects Design: trial

a.

Exact statisticb.

The statistic is an upper bound on F that yields a lower bound on the significance level.c.

Page 6

Page 7: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 7/29

Mauchly's Test of Sphericitya

Measure: sd_midslicewidth

Within Subjects Effect Mauchly's W

 Approx. Chi-

Square df Sig.

Epsilonb

Greenhouse-

Geisser trial .858 26.211 2 .000 .876

Mauchly's Test of Sphericitya

Measure: sd_midslicewidth

Within Subjects Effect

Epsilonb

Huynh-Feldt Lower-bound

trial .894 .500

Tests the null hypothesis that the error covariance matrix of the orthonormalized transformed dependent variables

is proportional to an identity matrix.

Design: Intercept + MovieCond

Within Subjects Design: trial

a.

May be used to adjust the degrees of freedom for the averaged tests of significance. Corrected tests are

displayed in the Tests of Within-Subjects Effects table.

b.

Tests of Within-Subjects Effects

Measure: sd_midslicewidth

Source

Type III Sum

of Squares df Mean Square F

trial Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

trial * MovieCond Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

Error(trial) Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

106.132 2 53.066 7.852

106.132 1.751 60.608 7.852

106.132 1.788 59.349 7.852

106.132 1.000 106.132 7.852

51.884 4 12.971 1.919

51.884 3.502 14.814 1.919

51.884 3.577 14.507 1.919

51.884 2.000 25.942 1.919

2324.761 344 6.758

2324.761 301.196 7.718

2324.761 307.581 7.558

2324.761 172.000 13.516

Page 7

Page 8: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 8/29

Tests of Within-Subjects Effects

Measure: sd_midslicewidth

Source Sig.

trial Sphericity AssumedGreenhouse-Geisser 

Huynh-Feldt

Lower-bound

trial * MovieCond Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

Error(trial) Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

.000

.001

.001

.006

.107

.116

.115

.150

Tests of Within-Subjects Contrasts

Measure: sd_midslicewidth

Source trial

Type III Sum

of Squares df Mean Square F Sig.

trial Linear  

Quadratic

trial * MovieCond Linear 

Quadratic

Error(trial) Linear  

Quadratic

81.759 1 81.759 9.013 .003

24.374 1 24.374 5.484 .020

41.355 2 20.677 2.279 .105

10.529 2 5.265 1.185 .308

1560.302 172 9.072

764.459 172 4.445

Tests of Between-Subjects Effects

Measure: sd_midslicewidth

Transformed Variable: Average

Source

Type III Sum

of Squares df Mean Square F Sig.

Intercept

MovieCond

Error 

12728.746 1 12728.746 1211.915 .000

6.966 2 3.483 .332 .718

1806.517 172 10.503

Estimated Marginal Means

Page 8

Page 9: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 9/29

1. MovieCond

Measure: sd_midslicewidth

MovieCond Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

.000000

1.000000

2.000000

4.787 .252 4.289 5.285

4.933 .236 4.468 5.399

5.075 .248 4.586 5.565

2. trial

Measure: sd_midslicewidth

trial Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

1

2

3

5.569 .223 5.128 6.010

4.627 .138 4.354 4.899

4.601 .262 4.083 5.118

3. MovieCond * t rial

Measure: sd_midslicewidth

MovieCond trial Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

.000000 1

2

3

1.000000 1

23

2.000000 1

2

3

5.496 .398 4.711 6.281

4.721 .246 4.236 5.206

4.145 .467 3.223 5.066

5.220 .372 4.487 5.954

4.380 .230 3.926 4.8335.200 .436 4.339 6.062

5.990 .391 5.219 6.761

4.780 .241 4.303 5.256

4.457 .459 3.551 5.362

Profile Plots

Page 9

Page 10: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 10/29

trial

321

   E  s   t   i  m  a   t  e   d   M  a  r  g   i  n  a   l   M  e  a  n  s

6.00

5.50

5.00

4.50

4.00

Estimated Marginal Means of sd_midslicewidth

2.000000

1.000000

.000000

MovieCond

DATASET ACTI VATE Dat aSet 9.

GLM Pre Post 1 t r i al 3_f ast BY Movi eCond

  / WSFACTOR=t r i al 3 Pol ynomi al

  / MEASURE=mi dsl i ceMT

  / METHOD=SSTYPE( 3)

  / PLOT=PROFI LE( t r i al *Movi eCond)

  / EMMEANS=TABLES( Movi eCond) COMPARE ADJ ( LSD)

  / EMMEANS=TABLES( t r i al ) COMPARE ADJ ( LSD)

  / EMMEANS=TABLES( Movi eCond*t r i al )

  / CRI TERI A=ALPHA( . 05)

  / WSDESI GN=t r i al

  / DESI GN=Movi eCond.

General Linear Model

[ DataSet9] J : \ Bi gRoomExper i ment s - J an- 22- 2014\ Ri t aLBOBr ad\ RMdata_meanMi dSl i c

eMT. sav

Page 10

Page 11: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 11/29

Within-Subjects

Factors

Measure: midsliceMT

trial

Dependent

Variable

1

2

3

Pre

Post1

trial3_fast

Between-Subjects Factors

N

MovieCond .000000

1.000000

2.000000

55

63

57

Multivariate Tests

a

Effect Value F Hypothesis df Error df Sig.

trial Pillai's Trace

Wilks' Lambda

Hotelling's Trace

Roy's Largest Root

trial * MovieCond Pillai's Trace

Wilks' Lambda

Hotelling's Trace

Roy's Largest Root

.453 70.749b

2.000 171.000 .000

.547 70.749b

2.000 171.000 .000

.827 70.749b

2.000 171.000 .000

.827 70.749b

2.000 171.000 .000

.019 .816 4.000 344.000 .515

.981 .813b

4.000 342.000 .517

.019 .810 4.000 340.000 .519

.016 1.388c

2.000 172.000 .252

Design: Intercept + MovieCondWithin Subjects Design: trial

a.

Exact statisticb.

The statistic is an upper bound on F that yields a lower bound on the significance level.c.

Mauchly's Test of Sphericitya

Measure: midsliceMT

Within Subjects Effect Mauchly's W

 Approx. Chi-

Square df Sig.

Epsilonb

Greenhouse-

Geisser 

trial .893 19.395 2 .000 .903

Page 11

Page 12: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 12/29

Mauchly's Test of Sphericitya

Measure: midsliceMT

Within Subjects Effect

Epsilonb

Huynh-Feldt Lower-boundtrial .923 .500

Tests the null hypothesis that the error covariance matrix of the orthonormalized transformed dependent variables

is proportional to an identity matrix.

Design: Intercept + MovieCond

Within Subjects Design: trial

a.

May be used to adjust the degrees of freedom for the averaged tests of significance. Corrected tests are

displayed in the Tests of Within-Subjects Effects table.

b.

Tests of Within-Subjects Effects

Measure: midsliceMT

Source

Type III Sum

of Squares df Mean Square F

trial Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

trial * MovieCond Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

Error(trial) Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

4204914.775 2 2102457.388 89.000

4204914.775 1.806 2327897.168 89.000

4204914.775 1.846 2278377.593 89.000

4204914.775 1.000 4204914.775 89.000

94958.261 4 23739.565 1.005

94958.261 3.613 26285.083 1.005

94958.261 3.691 25725.940 1.005

94958.261 2.000 47479.131 1.005

8126307.965 344 23622.988

8126307.965 310.686 26156.006

8126307.965 317.439 25599.609

8126307.965 172.000 47245.977

Page 12

Page 13: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 13/29

Tests of Within-Subjects Effects

Measure: midsliceMT

Source Sig.

trial Sphericity AssumedGreenhouse-Geisser 

Huynh-Feldt

Lower-bound

trial * MovieCond Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

Error(trial) Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

.000

.000

.000

.000

.405

.400

.401

.368

Tests of Within-Subjects Contrasts

Measure: midsliceMT

Source trial

Type III Sum

of Squares df Mean Square F Sig.

trial Linear  

Quadratic

trial * MovieCond Linear 

Quadratic

Error(trial) Linear  

Quadratic

4181816.386 1 4181816.386 140.552 .000

23098.390 1 23098.390 1.320 .252

80217.430 2 40108.715 1.348 .262

14740.831 2 7370.415 .421 .657

5117470.931 172 29752.738

3008837.034 172 17493.239

Tests of Between-Subjects Effects

Measure: midsliceMT

Transformed Variable: Average

Source

Type III Sum

of Squares df Mean Square F Sig.

Intercept

MovieCond

Error 

259145605.5 1 259145605.5 1807.122 .000

359804.587 2 179902.293 1.255 .288

24665206.78 172 143402.365

Estimated Marginal Means

1. MovieCond

Page 13

Page 14: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 14/29

Estimates

Measure: midsliceMT

MovieCond Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

.000000

1.000000

2.000000

729.434 29.481 671.244 787.624

668.390 27.545 614.020 722.761

713.382 28.959 656.222 770.542

Pairwise Comparisons

Measure: midsliceMT

(I) MovieCond (J) MovieCond

Mean

Difference (I-

J) Std. Error Sig.a

95% Confidence Interval for

Differencea

Lower Bound Upper Bound

.000000 1.000000

2.000000

1.000000 .000000

2.000000

2.000000 .000000

1.000000

61.044 40.347 .132 -18.595 140.682

16.052 41.325 .698 -65.517 97.620

-61.044 40.347 .132 -140.682 18.595

-44.992 39.967 .262 -123.881 33.897

-16.052 41.325 .698 -97.620 65.517

44.992 39.967 .262 -33.897 123.881

Based on estimated marginal means

 Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).a.

Univariate Tests

Measure: midsliceMT

Sum of

Squares df Mean Square F Sig.

Contrast

Error 

119934.862 2 59967.431 1.255 .288

8221735.592 172 47800.788

The F tests the effect of MovieCond. This test is based on the linearly

independent pairwise comparisons among the estimated marginal means.

2. trial

Estimates

Measure: midsliceMT

trial Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

1

2

3

817.921 21.766 774.959 860.884

694.339 18.368 658.084 730.595

598.946 16.790 565.805 632.086

Page 14

Page 15: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 15/29

Pairwise Comparisons

Measure: midsliceMT

(I) trial (J) trial

Mean

Difference (I-

J) Std. Error Sig.b

95% Confidence Interval for

Differenceb

Lower Bound Upper Bound

1 2

3

2 1

3

3 1

2

123.582*

16.810 .000 90.401 156.763

218.976*

18.470 .000 182.518 255.433

-123.582*

16.810 .000 -156.763 -90.401

95.394*

13.743 .000 68.267 122.521

-218.976*

18.470 .000 -255.433 -182.518

-95.394*

13.743 .000 -122.521 -68.267

Based on estimated marginal means

The mean difference is significant at the .05 level.*.

 Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).b.

Multivariate Tests

Value F Hypothesis df Error df Sig.

Pillai's trace

Wilks' lambda

Hotelling's trace

Roy's largest root

.453 70.749a

2.000 171.000 .000

.547 70.749a

2.000 171.000 .000

.827 70.749a

2.000 171.000 .000

.827 70.749a

2.000 171.000 .000

Each F tests the multivariate effect of trial. These tests are based on the linearly

independent pairwise comparisons among the estimated marginal means.

Exact statistica.

3. MovieCond * t rial

Measure: midsliceMT

MovieCond trial Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

.000000 1

2

3

1.000000 1

2

3

2.000000 1

2

3

836.801 38.761 760.293 913.310

728.998 32.710 664.433 793.563

622.502 29.899 563.485 681.519

806.526 36.216 735.041 878.012

649.705 30.563 589.378 710.031

548.939 27.937 493.797 604.082

810.436 38.075 735.282 885.590

704.315 32.131 640.893 767.737

625.395 29.370 567.423 683.368

Profile Plots

Page 15

Page 16: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 16/29

trial

321

   E  s   t   i  m  a   t  e   d   M  a  r  g   i  n  a   l   M  e  a  n  s

900.00

800.00

700.00

600.00

500.00

Estimated Marginal Means of midsliceMT

2.000000

1.000000

.000000

MovieCond

DATASET ACTI VATE Dat aSet 9.

SAVE OUTFI LE=' J : \ Bi gRoomExper i ment s - J an- 22- 2014\ Ri t aLBOBr ad\ RMdat a_meanMi dS

l i ceMT. sav'

  / COMPRESSED.

DATASET ACTI VATE Dat aSet 2.

DATASET CLOSE Dat aSet 9.

DATASET ACTI VATE Dat aSet 10.

GLM Pre Post 1 t r i al 3_f ast BY Movi eCond

  / WSFACTOR=t r i al 3 Pol ynomi al

  / MEASURE=sd_r

  / METHOD=SSTYPE( 3)

  / PLOT=PROFI LE( t r i al *Movi eCond)

  / EMMEANS=TABLES( Movi eCond) COMPARE ADJ ( LSD)

  / EMMEANS=TABLES( t r i al ) COMPARE ADJ ( LSD)

  / EMMEANS=TABLES( Movi eCond*t r i al )

  / CRI TERI A=ALPHA( . 05)

Page 16

Page 17: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 17/29

  / WSDESI GN=t r i al

  / DESI GN=Movi eCond.

General Linear Model

[ DataSet10] J : \ Bi gRoomExper i ment s - J an- 22- 2014\ Ri t aLBOBr ad\ RMdata_sdr . sav

Within-Subjects

Factors

Measure: sd_r 

trial

Dependent

Variable

1

2

3

Pre

Post1

trial3_fast

Between-Subjects Factors

N

MovieCond .000000

1.000000

2.000000

55

63

57

Multivariate Testsa

Effect Value F Hypothesis df Error df Sig.

trial Pillai's Trace

Wilks' Lambda

Hotelling's TraceRoy's Largest Root

trial * MovieCond Pillai's Trace

Wilks' Lambda

Hotelling's Trace

Roy's Largest Root

.041 3.670b

2.000 171.000 .028

.959 3.670b

2.000 171.000 .028

.043 3.670

b

2.000 171.000 .028.043 3.670

b2.000 171.000 .028

.053 2.359 4.000 344.000 .053

.947 2.377b

4.000 342.000 .052

.056 2.396 4.000 340.000 .050

.056 4.824c

2.000 172.000 .009

Design: Intercept + MovieCond

Within Subjects Design: trial

a.

Exact statisticb.

The statistic is an upper bound on F that yields a lower bound on the significance level.c.

Page 17

Page 18: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 18/29

Mauchly's Test of Sphericitya

Measure: sd_r 

Within Subjects Effect Mauchly's W

 Approx. Chi-

Square df Sig.

Epsilonb

Greenhouse-

Geisser trial .907 16.672 2 .000 .915

Mauchly's Test of Sphericitya

Measure: sd_r 

Within Subjects Effect

Epsilonb

Huynh-Feldt Lower-bound

trial .935 .500

Tests the null hypothesis that the error covariance matrix of the orthonormalized transformed dependent variables

is proportional to an identity matrix.

Design: Intercept + MovieCond

Within Subjects Design: trial

a.

May be used to adjust the degrees of freedom for the averaged tests of significance. Corrected tests are

displayed in the Tests of Within-Subjects Effects table.

b.

Tests of Within-Subjects Effects

Measure: sd_r 

Source

Type III Sum

of Squares df Mean Square F

trial Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

trial * MovieCond Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

Error(trial) Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

81.366 2 40.683 3.717

81.366 1.830 44.462 3.717

81.366 1.870 43.507 3.717

81.366 1.000 81.366 3.717

108.831 4 27.208 2.486

108.831 3.660 29.735 2.486

108.831 3.740 29.096 2.486

108.831 2.000 54.415 2.486

3765.219 344 10.945

3765.219 314.760 11.962

3765.219 321.672 11.705

3765.219 172.000 21.891

Page 18

Page 19: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 19/29

Tests of Within-Subjects Effects

Measure: sd_r 

Source Sig.

trial Sphericity AssumedGreenhouse-Geisser 

Huynh-Feldt

Lower-bound

trial * MovieCond Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

Error(trial) Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

.025

.029

.028

.056

.043

.049

.047

.086

Tests of Within-Subjects Contrasts

Measure: sd_r 

Source trial

Type III Sum

of Squares df Mean Square F Sig.

trial Linear  

Quadratic

trial * MovieCond Linear 

Quadratic

Error(trial) Linear  

Quadratic

77.430 1 77.430 5.831 .017

3.936 1 3.936 .457 .500

33.227 2 16.614 1.251 .289

75.604 2 37.802 4.390 .014

2284.122 172 13.280

1481.097 172 8.611

Tests of Between-Subjects Effects

Measure: sd_r 

Transformed Variable: Average

Source

Type III Sum

of Squares df Mean Square F Sig.

Intercept

MovieCond

Error 

20782.371 1 20782.371 327.291 .000

300.008 2 150.004 2.362 .097

10921.685 172 63.498

Estimated Marginal Means

1. MovieCond

Page 19

Page 20: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 20/29

Estimates

Measure: sd_r 

MovieCond Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

.000000

1.000000

2.000000

6.115 .620 4.891 7.340

5.491 .580 4.347 6.635

7.300 .609 6.097 8.503

Pairwise Comparisons

Measure: sd_r 

(I) MovieCond (J) MovieCond

Mean

Difference (I-

J) Std. Error Sig.b

95% Confidence Interval for

Differenceb

Lower Bound Upper Bound

.000000 1.000000

2.000000

1.000000 .000000

2.000000

2.000000 .000000

1.000000

.624 .849 .463 -1.051 2.300

-1.185 .870 .175 -2.901 .532

-.624 .849 .463 -2.300 1.051

-1.809*

.841 .033 -3.469 -.149

1.185 .870 .175 -.532 2.901

1.809*

.841 .033 .149 3.469

Based on estimated marginal means

The mean difference is significant at the .05 level.*.

 Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).b.

Univariate Tests

Measure: sd_r 

Sum of

Squares df Mean Square F Sig.

Contrast

Error 

100.003 2 50.001 2.362 .097

3640.562 172 21.166

The F tests the effect of MovieCond. This test is based on the linearly

independent pairwise comparisons among the estimated marginal means.

2. trial

Page 20

Page 21: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 21/29

Estimates

Measure: sd_r 

trial Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

1

2

3

6.712 .406 5.911 7.513

6.425 .410 5.615 7.235

5.770 .395 4.989 6.550

Pairwise Comparisons

Measure: sd_r 

(I) trial (J) trial

Mean

Difference (I-

J) Std. Error Sig.b

95% Confidence Interval for

Differenceb

Lower Bound Upper Bound

1 2

3

2 1

3

3 1

2

.287 .368 .437 -.440 1.014

.942*

.390 .017 .172 1.712

-.287 .368 .437 -1.014 .440

.655*

.297 .029 .068 1.242

-.942*

.390 .017 -1.712 -.172

-.655*

.297 .029 -1.242 -.068

Based on estimated marginal means

The mean difference is significant at the .05 level.*.

 Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).b.

Multivariate Tests

Value F Hypothesis df Error df Sig.

Pillai's trace

Wilks' lambda

Hotelling's trace

Roy's largest root

.041 3.670a

2.000 171.000 .028

.959 3.670a

2.000 171.000 .028

.043 3.670a

2.000 171.000 .028

.043 3.670a

2.000 171.000 .028

Each F tests the multivariate effect of trial. These tests are based on the linearly

independent pairwise comparisons among the estimated marginal means.

Exact statistica.

Page 21

Page 22: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 22/29

3. MovieCond * t rial

Measure: sd_r 

MovieCond trial Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

.000000 1

2

3

1.000000 1

2

3

2.000000 1

2

3

6.992 .723 5.565 8.419

5.851 .731 4.408 7.294

5.503 .704 4.113 6.893

6.252 .675 4.919 7.585

5.237 .683 3.889 6.585

4.984 .658 3.686 6.283

6.892 .710 5.490 8.293

8.186 .718 6.769 9.604

6.822 .692 5.457 8.187

Profile Plots

trial

321

   E  s   t   i  m  a   t  e   d   M  a  r  g   i  n  a   l

   M  e  a  n  s

8.00

7.00

6.00

5.00

Estimated Marginal Means of sd_r 

2.000000

1.000000

.000000

MovieCond

Page 22

Page 23: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 23/29

DATASET ACTI VATE Dat aSet 3.

GLM Pre Post 1 t r i al 3_f ast BY Movi eCond

  / WSFACTOR=t r i al 3 Pol ynomi al

  / MEASURE=si gnedMi dsl i ceEr r or

  / METHOD=SSTYPE( 3)

  / PLOT=PROFI LE( t r i al *Movi eCond)

  / EMMEANS=TABLES( Movi eCond) COMPARE ADJ ( LSD)

  / EMMEANS=TABLES( t r i al ) COMPARE ADJ ( LSD)

  / EMMEANS=TABLES( Movi eCond*t r i al )

  / CRI TERI A=ALPHA( . 05)

  / WSDESI GN=t r i al

  / DESI GN=Movi eCond.

General Linear Model

[ DataSet3] J : \ Bi gRoomExper i ment s - J an- 22- 2014\ Ri t aLBOBr ad\ RMdata_si gnmi dsl cw

dt h. sav

Within-Subjects

Factors

Measure:

signedMidsliceError 

trial

Dependent

Variable

1

2

3

Pre

Post1

trial3_fast

Between-Subjects Factors

N

MovieCond .000000

1.000000

2.000000

55

63

57

Page 23

Page 24: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 24/29

Page 25: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 25/29

Tests of Within-Subjects Effects

Measure: signedMidsliceError 

Source

Type III Sum

of Squares df Mean Square F

trial Sphericity AssumedGreenhouse-Geisser 

Huynh-Feldt

Lower-bound

trial * MovieCond Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

Error(trial) Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

103.564 2 51.782 17.359103.564 1.930 53.658 17.359

103.564 1.974 52.457 17.359

103.564 1.000 103.564 17.359

16.045 4 4.011 1.345

16.045 3.860 4.157 1.345

16.045 3.949 4.063 1.345

16.045 2.000 8.022 1.345

1026.123 344 2.983

1026.123 331.971 3.091

1026.123 339.571 3.022

1026.123 172.000 5.966

Tests of Within-Subjects Effects

Measure: signedMidsliceError 

Source Sig.

trial Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

trial * MovieCond Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

Error(trial) Sphericity Assumed

Greenhouse-Geisser 

Huynh-Feldt

Lower-bound

.000

.000

.000

.000

.253

.254

.253

.263

Page 25

Page 26: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 26/29

Tests of Within-Subjects Contrasts

Measure: signedMidsliceError 

Source trial

Type III Sum

of Squares df Mean Square F Sig.

trial Linear  Quadratic

trial * MovieCond Linear 

Quadratic

Error(trial) Linear  

Quadratic

82.115 1 82.115 23.266 .00021.449 1 21.449 8.804 .003

10.453 2 5.226 1.481 .230

5.592 2 2.796 1.148 .320

607.059 172 3.529

419.063 172 2.436

Tests of Between-Subjects Effects

Measure: signedMidsliceError 

Transformed Variable: Average

Source

Type III Sum

of Squares df Mean Square F Sig.Intercept

MovieCond

Error 

638.721 1 638.721 60.018 .000

7.747 2 3.873 .364 .695

1830.459 172 10.642

Estimated Marginal Means

1. MovieCond

Estimates

Measure: signedMidsliceError 

MovieCond Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

.000000

1.000000

2.000000

.932 .254 .430 1.433

1.170 .237 .701 1.638

1.213 .249 .721 1.706

Page 26

Page 27: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 27/29

Page 28: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 28/29

Pairwise Comparisons

Measure: signedMidsliceError 

(I) trial (J) trial

Mean

Difference (I-

J) Std. Error Sig.b

95% Confidence Interval for

Differenceb

Lower Bound Upper Bound

1 2

3

2 1

3

3 1

2

.915*

.172 .000 .575 1.254

.970*

.201 .000 .573 1.367

-.915*

.172 .000 -1.254 -.575

.056 .181 .758 -.301 .412

-.970*

.201 .000 -1.367 -.573

-.056 .181 .758 -.412 .301

Based on estimated marginal means

The mean difference is significant at the .05 level.*.

 Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).b.

Multivariate Tests

Value F Hypothesis df Error df Sig.

Pillai's trace

Wilks' lambda

Hotelling's trace

Roy's largest root

.164 16.736a

2.000 171.000 .000

.836 16.736a

2.000 171.000 .000

.196 16.736a

2.000 171.000 .000

.196 16.736a

2.000 171.000 .000

Each F tests the multivariate effect of trial. These tests are based on the linearly

independent pairwise comparisons among the estimated marginal means.

Exact statistica.

3. MovieCond * t rial

Measure: signedMidsliceError 

MovieCond trial Mean Std. Error  

95% Confidence Interval

Lower Bound Upper Bound

.000000 1

2

3

1.000000 1

2

3

2.000000 1

2

3

1.682 .337 1.017 2.347

.847 .285 .285 1.410

.265 .328 -.381 .912

1.672 .315 1.051 2.293

.740 .266 .214 1.265

1.097 .306 .493 1.701

1.845 .331 1.192 2.498

.869 .280 .316 1.421

.926 .322 .291 1.561

Profile Plots

Page 28

Page 29: Anova w High Speed Trial

7/21/2019 Anova w High Speed Trial

http://slidepdf.com/reader/full/anova-w-high-speed-trial 29/29

trial

321

   E  s   t   i  m  a   t  e   d   M  a  r  g   i  n  a   l   M  e  a  n  s

2.00

1.50

1.00

.50

.00

Estimated Marginal Means of signedMidsliceError 

2.000000

1.000000

.000000

MovieCond

Page 29